I. Ethics, Positive or Normative

In maintaining that human behaviour is naturally conditioned, that there is no difference between the kind of causality which conduct exhibits and that exhibited in any other case, determinism implies that the distinctions commonly drawn between normative and positive science, and between freedom and necessity, are unreal. This is not to say, as is sometimes supposed, either that we can have such a knowledge of a person's character and environment as will enable us to predict his whole history, or that his environment alone determines what his history will be. It would be absurd to make any such claim in regard to minds, when we find, in dealing with other things, that both character and environment have to be taken into account — indeed, if this were not so, we should have in turn to consider the environment, and not the character, of the environing things, and so on indefinitely — and that, in the investigation of both, new and unexpected factors are continually being revealed. Such discovery, however, is possible only if we can say that in certain situations things of a particular sort behave in a certain way; and this kind of description, which implies neither “freedom” nor subordination to “standards”, is the only way of expressing knowledge of human or any other behaviour. It is no more possible to uphold a theory of different kinds of causality than consistently to believe in different kinds of truth or reality.

Those who distinguish positive from normative science contend that, while the former can tell us “what is”, it cannot tell us “what ought to be”. In order to know that, we require to invoke “norms”; standards or ideals to which things may conform more or less closely, but which are in some sense above these things. Such norms, it is said, appear in practice, where we must have ends to which our actions are directed. And while natural science can tell us as much about the connections of things as will enable us to select means to ends, it cannot determine the ends themselves. It supplies us, not with the absolute or categorical commands of morality, but only with hypothetical commands. Medical science, for example, can say that if we desire health, we should not deprive ourselves of fresh air, but we have to go beyond medical science to find out whether health ought to be desired. Now action which is determined by pursuit of an absolute end is thereby subject to a different sort of causality from that which prevails among things which have not ends; it is “free” action.

  ― 215 ―
The illogicality of this theory appears at once from its conception of the different sorts of reality which attach to norms and to the things which come under these norms, or from its attempt to distinguish values from facts. If the statement that something “ought to be” has any meaning, it can only be that the thing is, positively, obligatory; that this is a matter of fact. When such a statement is taken to be true, it can be dealt with by means of the ordinary logical mechanism of assertion and denial, proof and testing of hypotheses, definition and division — and in no other way. If we accept the term “obligatory”, then we shall say that certain things are or are not obligatory, just as we may say that they are or are not red. But if we set out to show that they “are” in different senses, we can do so only by using propositions in relation to which an exactly similar operation will be required. The consequent impossibility of making any definite assertion would be fatal to any science, “normative” or positive.

To get over this difficulty it has been suggested that all science is really normative, that “physical science” is only that part of Science which deals with means, and is thus a mere abstraction from the fundamental consideration of ends. But how is this abstraction possible? Suppose it were granted that medical science deals with means (a fairly plausible assumption), what are they, and how can it deal with them so as to be of any use to us in conduct? They must have characters of their own, they must exist, or they cannot be studied. And, however strenuously it may be asserted that in this existence they are only “relative”, it must be admitted that they absolutely lead to the desired ends, if we are to get any advantage from knowing them. And, finally, this end that they lead to must have certain physical characteristics, if physical science is to tell us how to get to it. It appears, then, that medical and ethical science may consider the same things, but will concentrate on different features of them. This is a commonplace of scientific investigation, and does not justify the attempt to take those features, in which ethics is specially interested, apart from the rest, and as an end to which the rest are a means. Incidentally, we should then have two sorts of “means to end” relation; fresh air would promote health in quite a different way from that in which health promoted goodness. Yet both relations would be determined in exactly the same way — by seeing what followed from certain conditions.

To avoid these inconsistencies we must maintain that it is possible to find physical occurrences without moral characteristics, and also to find physical occurrences with moral characteristics — which, of course, need have nothing to do with ends. If the value of anything were something above its occurrence, it would be unaffected by whether the thing occurred or not, and thus the occurrence itself would have no value, i.e., the value would not, strictly speaking, be of anything, and, being quite apart from events, certainly could not be chosen as an end. On the other hand if there is anything in the occurrence itself which could be regarded as its “conformity to a standard”, then this character is what we mean by the thing's value and we do not require to look beyond the thing itself

  ― 216 ―
in order to know an ethical truth. The alternative is that value is a mere label which might be arbitrarily attached to anything.

It will, of course, be said that nothing ever happens which quite comes up to our ideals. But, in the first place, this comparison could not be made unless our ideals were at least thought of as happening, i.e., as natural. And while we can and do think of goods which may occur in the future, the same applies to the subject-matter of any science. There is nothing logically peculiar about the future; what we think of as existing then, we think of as existing in the same sense as what exists now. And the characters we attribute to future occurrences are those which we have already found in similar occurrences. Allowing, then, that we can suppose that something better will exist than anything that has so far existed, we conceive its goodness to be of the same sort as that of things already in existence. But, in the second place, suppositions of this kind are for the most part false and useless; the imagined situation would very often not be particularly good; and if we say “better but not possible”, either we mean that it is not possible here but has occurred somewhere else and been observed to be better, or we are speaking about the non-existent and cannot significantly call “it” better or worse than what exists.